Long-term total parenteral nutrition-induced hepatobiliary dysfunction in a rabbit model.

BACKGROUND/PURPOSE: Currently, the reason for hepatobiliary dysfunction associated with long-term total parenteral nutrition (TPN) is much debated and still unclear. No agreement can be achieved about whether bacteriotoxins and sepsis, enteral starvation, consequences of abdominal operations, or the TPN solution itself is the real cause for the disease. Animal models were criticized for their short period of TPN and their failure to demonstrate cholestasis and bile duct proliferation. The aim of this study was to establish an animal model for long-term TPN in which the same alterations of the hepatobiliary system as observed in humans could be produced. METHODS: In this model, rabbits could be kept for the first time under continuous TPN for 4 weeks. Three serial liver biopsy sections were taken operatively from each animal and biochemical analyses were performed four times. A control group of enterally fed rabbits underwent exactly the same procedure in respect to operations and handling, so that differences in macroscopical, biochemical, and histological changes between both groups could be attributed exclusively to TPN. RESULTS: Only in the TPN group gallbladder distension developed in all animals after 1 week. After 3 and 4 weeks, viscous dark bile, sludge and stones, a slight rise in direct bilirubin, and a decline in plasma albumin and alkaline phosphatase was noted. In both groups liver biopsy results showed a similar degree of mild portal inflammation and single-cell necrosis at equivalent time points. These changes could be caused by antiseptics, antibiotics, anesthesia, and operations. Although mild to moderate proliferative changes and no hydropic degeneration developed in the control group during the same time, the TPN group generated marked proliferative and degenerative changes. We noted as early as 1 week after starting TPN a severe hydropic degeneration in 90% of the animals. Fibrosis and bile duct proliferation increased from a slight degree after 1 week up to a moderate to severe degree after 3 and 4 weeks, respectively. CONCLUSIONS: The hepatobiliary alterations associated with TPN in children, which cannot be separated clinically from consequences of multiple other factors, can almost identically be reproduced in our rabbit model as a clear consequence of TPN. Furthermore, the hydropic degeneration of the liver cells begins in zone 3 and is an early predominant feature of hepatobiliary dysfunction in rabbits and infants. It must be rated as a response to a direct cytotoxic effect on the liver cell.